Is there no finer resolution than genus for Eucalyptus species? It is surprising to see such low-resolution categories in a paper about phylogenetic diversity.

The McKinnon et al. reference appears to have species-level resolution.

In fact, what we showed was that FD and PD are essentially uncorrelated... after removing the effect of species richness! Both are strongly correlated with S by the math of the calculation. But the general point you make stands.